Improving construction site safety through leader-based verbal safety communication

Abstract

Background

The construction industry is one of the most injury-prone industries, in which production is usually prioritized over safety in daily on-site communication. Workers have an informal and oral culture of risk, in which safety is rarely openly expressed. This paper tests the effect of increasing leader-based on-site verbal safety communication on the level of safety and safety climate at construction sites. Method: A pre-post intervention-control design with five construction work gangs is carried out. Foremen in two intervention groups are coached and given bi-weekly feedback about their daily verbal safety communications with their workers. Foremen-worker verbal safety exchanges (experience sampling method, n = 1,693 interviews), construction site safety level (correct vs. incorrect, n = 22,077 single observations), and safety climate (seven dimensions, n = 105 questionnaires) are measured over a period of up to 42 weeks. Results: Baseline measurements in the two intervention and three control groups reveal that foremen speak with their workers several times a day. Workers perceive safety as part of their verbal communication with their foremen in only 6-16% of exchanges, and the levels of safety at the sites range from 70-87% (correct observations). Measurements from baseline to follow-up in the two intervention groups reveal that safety communication between foremen and workers increases significantly in one of the groups (factor 7.1 increase), and a significant yet smaller increase is found when the two intervention groups are combined (factor 4.6). Significant increases in the level of safety are seen in both intervention groups (7% and 12% increases, respectively), particularly in regards to 'access ways' and 'railings and coverings' (39% and 84% increases, respectively). Increases in safety climate are seen in only one of the intervention groups with respect to their 'attention to safety.' No significant trend changes are seen in the three control groups on any of the three measures. Conclusions: Coaching construction site foremen to include safety in their daily verbal exchanges with workers has a significantly positive and lasting effect on the level of safety, which is a proximal estimate for work-related accidents. It is recommended that future studies include coaching and feedback at all organizational levels and for all involved parties in the construction process. Building client regulations could assign the task of coaching to the client appointed safety coordinators or a manager/supervisor, and studies should measure longitudinal effects of coaching by following foremen and their work gangs from site to site.

Introduction

The construction industry is one of the most injury-prone industries worldwide in terms of serious injuries, lost work time (Brenner & Ahern, 2000, Derr et al., 2001, Dong et al., 1995, Kartam et al., 2000, Kines, 2001, Larsson & Field, 2002, Loosemore & Andonakis, 2007, Wadick, 2007), hospitalization (Spangenberg, Hannerz, & Tüchsen, 2005), disability (Hannerz, Spangenberg, & Tüchsen, 2005), and mortality (Tüchsen, Hannerz, & Spangenberg, 2005), and there is a great need to improve worker safety at construction sites (Spangenberg, Mikkelsen, et al., 2005, Spangenberg et al., 2003, Spangenberg et al., 2002). The majority of construction enterprises have a reactive approach to risk management – waiting until after accidents or injunctions occur, rather than taking a proactive approach (Mikkelsen, Dyreborg, & Spangenberg, 2003). The use of injury incidence rates as the sole safety indicator in companies and at their construction sites is statistically inappropriate, as these rates must often operate on an industry or national level in order to show statistically significant changes over time (Spangenberg et al., 2002). There is a need to supplement negative and reactive feedback/measures (accident statistics) with positive and proactive feedback/feedback (safety statistics).

Many proactive safety observation techniques have been developed over the last few decades, at least two of which have shown to be predictive of accidents. Laitinen, Marjamäki, and Päivärinta's (1999) safety index provides a proactive measure of the physical and behavioral safety level at a work site (percent correct vs. incorrect safety observations), as do ‘traffic light’ observation systems that rate safety conditions/behavior as red - high-risk, yellow – medium-risk, green – low-risk (Mikkelsen, Spangenberg, & Kines, 2010). In spite of these positive results, there is still uncertainty as to how to effectively motivate and provide positive feedback at construction worksites.

A number of studies provide evidence showing that the aspects management prioritize have a spill-over effect on workers’ attitudes and behavior (Zohar, 2008). Studies regarding leader-worker exchanges and safety communication, safety climate, and safety citizenship have used questionnaires focused on social exchange theory and measuring the quality of exchanges (Hofmann et al., 2003, Hofmann & Morgeson, 1999). They found that leader-worker exchanges tended to stimulate employees to raise safety issues. A meta-analysis performed by Stajkovic and Luthans (2003) indicated that supervisory feedback and recognition were amongst the most powerful incentives influencing job performance. Daily supervisory feedback regarding safe and unsafe behavior and conditions provides an indication of the true priorities between production and safety, particularly when work is performed under extreme time pressure. Such data led Zohar and colleagues to develop a proactive leader-based safety intervention method focusing on the modification of supervisory monitoring and rewarding practices, by encouraging supervisors to express high safety priority during daily informal exchanges with workers (Zohar, 2002, Zohar & Luria, 2003).

Their method was based on quasi-random sampling of on-site supervisory interactions, using the exchange recipients (i.e., workers) as the pertinent data sources. Workers were approached at random times during the workday by a member of the intervention team asking them to report about their most recent exchange with their supervisor. Supervisors received bi-weekly feedback and coaching concerning the frequency of their safety related communication with subordinates. The supervisors used the data to self-monitor progress toward designated improvement goals. The intervention is leader based, rather than worker based, and the information collected is primarily used to motivate change in supervisory/managerial safety practices, and secondarily and indirectly to motivate change in worker safety practices. This puts focus on possible contributing accident factors both proximal and distal to adverse events (accidents/injuries), allowing managers and supervisors to positively influence the worksite safety climate in combination with coordinating the production processes. As opposed to behavior-based interventions targeted solely at the worker level, this leader-based approach recognizes that contributing factors to adverse events are multifaceted and complex, requiring conditions allowing for true company (multi-level) priority of safety. In this way the leader-based intervention is both more feasible and effective as the variables that influence worker behavior (e.g., supervisor exchanges) are placed under tighter control. The above studies have been carried out in manufacturing, food processing, and the military (Zohar, 2002, Zohar & Luria, 2003), but yet have to be tested in the construction industry.

Construction sites are hierarchically organized with site managers, foremen, and construction workers. The construction industry is characterized by traditional masculine values such as freedom, independence, resourcefulness, and toughness, with an often informal and oral culture of risk in which safety knowledge is tacitly understood without being openly expressed (Wadick, 2007, Baarts, 2004). Professional training of construction workers usually consists of practical on-site experiences. As much of the training is based on hands-on experiences, work practices are based on traditions, often passed down through foremen and more experienced coworkers, who in turn also are self-taught. Knowledge of and attitudes toward safety often vary from worker to worker, and safety practices, norms, and attitudes are continuously negotiated between workers and foremen (Baarts, 2004). Yet changing construction workers’ attitudes regarding safety is a constant challenge. Enforcement strategies often do not take into consideration the construction industry's culture or social structure, and work environment officials are often met with a form of passive resistance (Wadick, 2007).

A foreman's attitude to safety has great significance, and they are singled out as a crucial element in endeavors to enhance safe behavior at construction sites. A recent study provides evidence that foremen/supervisors have a stronger influence on construction workers’ safety attitudes than the workers’ colleagues (Dingsdag, Biggs, & Sheahan, 2008). However, safe behavior is often inhibited at construction sites as production goals are often prioritized over safety goals (Wadick, 2007). The construction industry is under constant pressure to reduce costs and at the same time improve quality. This is particularly challenging with rising costs in labor and materials, and in building increasingly complex structures (Errasti et al., 2007, Olsson, 1998). Each construction project is different due to varying project designs, a transient and multinational work force, multitudinous trades, and coordination of large amounts of equipment and materials. Coordination and cooperation at construction sites is a constant challenge, where the work of the different trades and subcontractors is interdependent and in some cases reciprocally and sequentially interdependent (Kadefors, 1995).

Amongst construction workers in particular, behavior may be guided by the principle of maximizing expected utility, whereby the benefits of unsafe behavior often appear to outweigh those of safe behavior (Zohar & Luria, 2003). Slower pace may lead to lower wages, especially for construction workers who often work at piece rates and various bonus systems. Safe behavior most often results in non-events, and unsafe behavior appears in most cases to lead to tangible benefits (Ibid; Reynolds & Schiffbauer, 2004). Zohar and Luria (2003) propose that safe behavior may be enhanced by introducing short-term rewards that outweigh immediate costs. Social incentives delivered by superiors, such as recognition and personal attention, have been shown to be as least as effective as financial incentives (Ibid). In addition, Mitchell (2000) found that quick and continuous feedback regarding safety is particularly important in temporary work places, such as in the construction industry.

This study sought to apply coaching and feedback of foremen/supervisors in an, as of yet, unexplored sector, the construction industry. Coaching and feedback of construction site foremen regarding daily verbal safety exchange with workers was hypothesized to result in: (a) increased frequency of verbal safety exchanges between foremen and workers, (b) increased levels of workers’ safe behavior and physical site safety, and (c) a heightened safety climate.